blob: edf4cda6a507f0c430da5db161fbd2d375fe9304 [file]
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py UTC_ARGS: --version 6
; This is the loop in c++ being vectorize in this file with
;vector.reverse
; #pragma clang loop vectorize_width(4, scalable)
; for (int i = N-1; i >= 0; --i)
; a[i] = b[i] + 1.0;
; RUN: opt -passes=loop-vectorize -mtriple=riscv64 -mattr=+v \
; RUN: -vplan-print-after=printOptimizedVPlan -force-vector-width=4 -scalable-vectorization=on \
; RUN: -disable-output < %s 2>&1 | FileCheck %s
define void @vector_reverse_i64(ptr nocapture noundef writeonly %A, ptr nocapture noundef readonly %B, i32 noundef signext %n) {
; CHECK-LABEL: VPlan for loop in 'vector_reverse_i64'
; CHECK: VPlan 'Initial VPlan for VF={vscale x 4},UF={1}' {
; CHECK-NEXT: Live-in vp<[[VP0:%[0-9]+]]> = VF * UF
; CHECK-NEXT: Live-in vp<[[VP1:%[0-9]+]]> = vector-trip-count
; CHECK-NEXT: vp<[[VP2:%[0-9]+]]> = original trip-count
; CHECK-EMPTY:
; CHECK-NEXT: ir-bb<entry>:
; CHECK-NEXT: EMIT vp<[[VP2]]> = EXPAND SCEV (1 + (-1 * (1 umin %n))<nuw><nsw> + %n)
; CHECK-NEXT: Successor(s): scalar.ph, vector.ph
; CHECK-EMPTY:
; CHECK-NEXT: vector.ph:
; CHECK-NEXT: Successor(s): vector loop
; CHECK-EMPTY:
; CHECK-NEXT: <x1> vector loop: {
; CHECK-NEXT: vp<[[VP3:%[0-9]+]]> = CANONICAL-IV
; CHECK-EMPTY:
; CHECK-NEXT: vector.body:
; CHECK-NEXT: CURRENT-ITERATION-PHI vp<[[VP5:%[0-9]+]]> = phi ir<0>, vp<%current.iteration.next>
; CHECK-NEXT: EMIT-SCALAR vp<%avl> = phi [ vp<[[VP2]]>, vector.ph ], [ vp<%avl.next>, vector.body ]
; CHECK-NEXT: EMIT-SCALAR vp<%evl> = EXPLICIT-VECTOR-LENGTH vp<%avl>
; CHECK-NEXT: vp<[[VP6:%[0-9]+]]> = DERIVED-IV ir<%n> + vp<[[VP5]]> * ir<-1>
; CHECK-NEXT: vp<[[VP7:%[0-9]+]]> = SCALAR-STEPS vp<[[VP6]]>, ir<-1>, vp<%evl>
; CHECK-NEXT: CLONE ir<%i.0> = add nsw vp<[[VP7]]>, ir<-1>
; CHECK-NEXT: EMIT-SCALAR ir<%idxprom> = zext ir<%i.0> to i64
; CHECK-NEXT: CLONE ir<%arrayidx> = getelementptr inbounds ir<%B>, ir<%idxprom>
; CHECK-NEXT: vp<[[VP8:%[0-9]+]]> = vector-end-pointer i32, ir<%arrayidx>, vp<%evl>
; CHECK-NEXT: WIDEN ir<%0> = vp.load vp<[[VP8]]>, vp<%evl>
; CHECK-NEXT: WIDEN ir<%add9> = add ir<%0>, ir<1>
; CHECK-NEXT: CLONE ir<%arrayidx3> = getelementptr inbounds ir<%A>, ir<%idxprom>
; CHECK-NEXT: vp<[[VP9:%[0-9]+]]> = vector-end-pointer i32, ir<%arrayidx3>, vp<%evl>
; CHECK-NEXT: WIDEN vp.store vp<[[VP9]]>, ir<%add9>, vp<%evl>
; CHECK-NEXT: EMIT vp<%current.iteration.next> = add vp<%evl>, vp<[[VP5]]>
; CHECK-NEXT: EMIT vp<%avl.next> = sub nuw vp<%avl>, vp<%evl>
; CHECK-NEXT: EMIT vp<%index.next> = add vp<[[VP3]]>, vp<[[VP0]]>
; CHECK-NEXT: EMIT branch-on-count vp<%index.next>, vp<[[VP1]]>
; CHECK-NEXT: No successors
; CHECK-NEXT: }
; CHECK-NEXT: Successor(s): middle.block
; CHECK-EMPTY:
; CHECK-NEXT: middle.block:
; CHECK-NEXT: Successor(s): ir-bb<for.cond.cleanup>
; CHECK-EMPTY:
; CHECK-NEXT: ir-bb<for.cond.cleanup>:
; CHECK-NEXT: No successors
; CHECK-EMPTY:
; CHECK-NEXT: scalar.ph:
; CHECK-NEXT: Successor(s): ir-bb<for.body>
; CHECK-EMPTY:
; CHECK-NEXT: ir-bb<for.body>:
; CHECK-NEXT: IR %indvars.iv = phi i32 [ %n, %entry ], [ %indvars.iv.next, %for.body ] (extra operand: ir<%n> from scalar.ph)
; CHECK-NEXT: IR %i.0.in8 = phi i32 [ %n, %entry ], [ %i.0, %for.body ] (extra operand: ir<%n> from scalar.ph)
; CHECK-NEXT: IR %i.0 = add nsw i32 %i.0.in8, -1
; CHECK-NEXT: IR %idxprom = zext i32 %i.0 to i64
; CHECK-NEXT: IR %arrayidx = getelementptr inbounds i32, ptr %B, i64 %idxprom
; CHECK-NEXT: IR %0 = load i32, ptr %arrayidx, align 4
; CHECK-NEXT: IR %add9 = add i32 %0, 1
; CHECK-NEXT: IR %arrayidx3 = getelementptr inbounds i32, ptr %A, i64 %idxprom
; CHECK-NEXT: IR store i32 %add9, ptr %arrayidx3, align 4
; CHECK-NEXT: IR %cmp = icmp ugt i32 %indvars.iv, 1
; CHECK-NEXT: IR %indvars.iv.next = add nsw i32 %indvars.iv, -1
; CHECK-NEXT: No successors
; CHECK-NEXT: }
;
entry:
br label %for.body
for.body:
%indvars.iv = phi i32 [ %n, %entry ], [ %indvars.iv.next, %for.body ]
%i.0.in8 = phi i32 [ %n, %entry ], [ %i.0, %for.body ]
%i.0 = add nsw i32 %i.0.in8, -1
%idxprom = zext i32 %i.0 to i64
%arrayidx = getelementptr inbounds i32, ptr %B, i64 %idxprom
%1 = load i32, ptr %arrayidx, align 4
%add9 = add i32 %1, 1
%arrayidx3 = getelementptr inbounds i32, ptr %A, i64 %idxprom
store i32 %add9, ptr %arrayidx3, align 4
%cmp = icmp ugt i32 %indvars.iv, 1
%indvars.iv.next = add nsw i32 %indvars.iv, -1
br i1 %cmp, label %for.body, label %for.cond.cleanup
for.cond.cleanup:
ret void
}